Artificial Intelligence

Fixing Internet Congestion Through Artificial Intelligence


Since first being introduced in the 1990s to the general public, the internet is getting simpler to use every day.

From access and availability to speed and user experience, we’ve come a long way from the dial-up internet we started out with. Consumers now have access to better quality media than previously available, with 4k video, clearer smartphone pictures, and better music quality.

People today are also using more data than ever before, with forecasts predicting the demand rising exponentially as the digital population grows.

With rising demand internet congestion will get worse and, for the first time since it was adopted by the general public, using the internet could become a less pleasurable experience than it currently is. As it is, the very infrastructure of the internet is struggling to keep up with the demand for content, leading to ever-increasing internet congestion.

I had a chance to speak with William Erbey, a serial entrepreneur who has founded six multi-billion-dollar public companies. Erbey is looking to solve the issue of congestion. With a background in informatics, Erbey is leveraging AI and emerging technology to build a scalable, sustainable solution.

A Dated Approach to Handling Internet Traffic

One of the most significant contributors to internet congestion is the way data is routed through the web. The vast majority of the internet uses a 40-year-old Border Gateway Protocol to route content through the shortest path possible. In theory, this approach would suggest very efficient data delivery, but it falls considerably short in practice.

What the Border Gateway protocol fails to consider is whether routers along the way are overloaded. The lack of communication between network hubs results in overlapped efforts, waste, and traffic congestion. Instead of an organized highway where information is sent through uncongested pathways, it is squeezed through just a handful of routers.

The most data-heavy online activity is video streaming, which accounts for 58% of total internet usage. A further 22% stems from torrents and 8% from gaming. It’s important to realize that video streaming services are growing almost 20% a year. When taking into account how 4k streaming is slowly becoming more mainstream, video streaming will only increase its data consumption.

Hardware-Based Solutions Have Diminishing Returns

How is this growing need being dealt with? Not well, according to Erbey. He points out that today’s solution to internet congestion isn’t to overhaul the way traffic is routed, but rather to add more hardware. This means placing more routers, creating bigger traffic hubs, and laying down more fiber.

While this approach enables internet service providers to advertise ever-faster connections, the consumer will rarely ever see video streaming or download speeds that match the promised number because a hardware backed solution is an approach with fast diminishing returns. It’s a temporary solution which isn’t viable in the long run as it’s not addressing the underlying issues.

Leveraging AI to Tackle Inefficiencies

Erbey asserts that leveraging AI and other emerging technologies is the way towards a sustainable and scalable solution to internet congestion. A company he invested in, System73, seeks to tackle internet congestion by combining AI and machine learning through Kuno Flow which is an AI-augmented, tree-based, peer-to-peer network that can enhance streaming quality and capacity.

“Kuno Stream implements an intelligent decision-making approach to directing internet traffic. It forecasts capacity and demand and identifies optimal delivery paths after weighing these variables. Since Kuno Stream deploys a managed overlay to the internet, it can help alleviate or even fix network issues by constructing non-congested data paths,” Erbey explained.

Since Kuno Stream is a peer-to-peer service, a significant amount of traffic can be offloaded to individual devices. This means devices can share content instead of requesting the same file multiple times.

“This approach results in an organized flow of information, where no single path becomes overloaded. Since there are more pathways due to the inclusion of individual devices, network capacity can expand exponentially to meet the growing demand for data,” says Erbey.

More Devices May Not Mean More Problems

Each new phone, tablet, or computer that joins PolyNet makes it quicker and more efficient. As opposed to BGP networking which where each new device adds to traffic congestion since more data needs to be squeezed through the same number of pathways.

The next time your Netflix stream slows to a crawl, it’s probably because your data is caught in the digital equivalent of a traffic jam. This is a common occurrence during peak hours, which explains why it may feel as if your download speeds are faster during certain times of the day. Until this issue is addressed by content based companies, the issue will only grow over time, since people are consuming more and more data each year. Improvements leading to 8k video, better smartphone photos, and higher quality music will also magnify the problem.

While internet congestion is a growing issue, it’s also starting to be addressed. Erbey is one of many entrepreneurs who are seeking to streamline inefficiencies in the internet’s data processing. By decentralizing the content delivery process, routing internet traffic intelligently, and forecasting demand, we can still fix the internet before it’s too late.

Raphael Badani is a geopolitical risk consultant and interactive simulation designer in the private sector.


© 2019 Newsmax Finance. All rights reserved.



READ SOURCE

This website uses cookies. By continuing to use this site, you accept our use of cookies.